Chapter 5: Six Causes and Four Conditions—Another Algorithm for Relations¶

The end of the last chapter left a distinction standing. Fascicle one hundred fifty-seven says that a characteristic and what it characterizes must share one fruit, that the characteristic accompanies what it characterizes and is not separate from it, and that there is no before or after between them; whereas possession and what is possessed do not share one fruit, do not of necessity accompany one another, are not inseparable, and may stand before or after. Two dharmas are tied together in both cases, and the two ties are not equally tight. To state that difference clearly, one attainment or one persistence is not enough; a whole doctrine of causes and conditions is wanted. The Sarvāstivāda has one, the Theravāda has one, and the two are shaped very differently. This chapter compares them.

The lists are not set out again. Part I has already given the definitions of the six causes, the four conditions, and the five fruits, and Chapter 5 of Part 14 gave the twenty-four conditions; this chapter does not do either a second time.⁴ What is asked here is something else: twenty-four items on one side, six plus four plus five on the other—where does that difference in number come from, and where does the cost of it fall.

Look first at the shape. The Theravāda's twenty-four are a flat list. Each item is a type of relation, none is subordinate to any other, and one may begin reading anywhere. The Sarvāstivāda's apparatus is a subsumption in two tiers: the four conditions comprise everything, the causal condition among them comprises five of the six causes, and the remaining efficient cause is distributed among the other three conditions. The levels are distinct and the items are fewer by more than half. It looks as though the same business had been tidied up.

But the tidiness was fitted afterwards. The four graphs "six causes and four conditions," used together, occur exactly once in the two hundred fascicles of the Mahāvibhāṣā: in fascicle twenty-three, and not in a discussion of causation at all, but in a definition of what makes an Abhidharma master—one must come to know the meaning of the Abhidharma basket through fourteen items, of which the six causes and four conditions are merely the first two; and the same passage records a variant that makes it seven items instead.¹ In the thirty fascicles of the Abhidharmakośa and the eighty of the Nyāyānusāra, the four graphs never occur together at all.²³ Each of the three lists exists on its own; what does not exist is the account that welds them into a single three-tier machine. That account is a convenience of later hands. Chapter 2 recorded that the table of seventy-five dharmas in five categories is not found in the Mahāvibhāṣā; Chapter 3 recorded that the four graphs "existence in the three times" are not found there; Chapter 4 recorded that prior possession and co-present possession are not found there. This is the fourth instance.

The treatise states its own provenance. Fascicle sixteen opens its discussion of the six causes with a blunt sentence: these six causes are not stated in the sūtras; the sūtras state only that there are four conditions; the six causes are stated here because one wishes to distinguish the conditions by way of cause.¹ The six causes were made in order to cut the four conditions of the sūtras finer, and the treatise knows it. What is interesting is the other end: fascicle one hundred twenty-one, discussing the five fruits, says that in the sūtras the fruits are stated to be five in number.¹ Of three lists, two claim scriptural warrant and the middle one admits to being new.

The weight of each of the three lists, together with the provenance the treatise claims for it:

Having admitted it, the treatise then tries to take it back. In the same fascicle sixteen someone next argues that the six causes were stated in scripture too, only that the scripture has been lost—the Ekottarāgama originally increased from one item up to a hundred and now retains only one to ten; the venerable Kātyāyanīputra and others, by the power of the knowledge born of resolve, beheld the place in the sūtras where the six causes were taught, and composed the treatise on that basis. It is added that on the day the arhat Śāṇavāsa entered final nirvāṇa, seventy-seven thousand birth-story scriptures and ten thousand Abhidharma treatises vanished from sight at once—if one master's passing carried off that much, how much more the hundreds and thousands of masters after him.¹ Someone else then says that although no single sūtra sets out the six in order, they lie scattered through the sūtras, and proceeds to match one scriptural sentence to each of the six.¹ This book takes the structure of that defence to be evidence in itself: when a doctrine has to be guaranteed by something already lost, the person saying so knows it is new.

There are also two views about how the subsumption runs. On the question which comprises which, cause or condition, fascicle sixteen gives two answers: on one, the two comprise each other according to the case; on the other, condition comprises cause and cause does not comprise condition, since the immediately antecedent condition and the objective condition fall outside the six causes altogether.¹ The difference is not small. On the second answer, two of the four conditions have no counterpart whatever among the causes, and the two tiers are not one nested inside the other but two figures that overlap. A more conspicuous passage stands nearby: fascicle sixteen says that the six causes were framed in order to display four fruits—the associated and co-existent causes display the fruit of human effort, the homogeneous and universal causes display the outflowing fruit, the cause of maturation displays the fruit of maturation, and the efficient cause displays the dominant fruit—and that one who has grasped this sees as clearly as one looking at an āmalaka fruit in the palm of the hand.¹ What is counted out there is four fruits, not five; the fruit of disjunction has no place on that table. Six, four, and five do not mesh with one another; the seams were smoothed over later.

Whether the apparatus was real was asked at the time. Fascicle fifty-five, on the four conditions, begins by recording the opposition: some hold that conditions have no real nature, as the Dārṣṭāntikas do. Their reason comes from a scriptural sentence—scripture says that ignorance is the condition of formations; but ignorance has no distinguishing characteristic of its own while formations do, and how should a dharma without a differentiating characteristic serve as condition to dharmas that have one and be itself really existent. The Bhadanta puts it more plainly: the masters posit the names of the conditions each according to his own notion, and there is no real nature to them.¹ The claim is that those condition-names are arranged as the framers of the doctrine saw fit, and that there is no such thing in the world as the immediately antecedent condition. The Mahāvibhāṣā's reply repays attention: if the nature of the conditions is not really existent, then no dharma at all has real nature, since the four conditions comprise all dharmas.¹ This is a reply that yields ground in order to gain it, and it yields too much—it concedes that a condition has no substance over and above the dharmas, that a condition is simply the dharmas themselves gone through once more by role. The end of this chapter returns to the point. The treatise then adds that if the nature of conditions were not really existent there would be no distinction of the three grades of enlightenment; and Ghoṣaka adds that if it were not, a teacher could not raise a pupil's understanding from poor to excellent, and a pupil could never become a teacher.¹ This book judges neither of the last two to be an argument at all: they say what would follow if there were no causal relation, whereas what the Dārṣṭāntikas deny is not causal relation but that the relation is a further substance.

It should also be set down whom the apparatus was aimed at. Fascicle sixteen says the six causes were framed to stop the doctrines of no cause and wrong cause: some outsiders hold that dharmas arise causelessly, others that there is an unequal cause.¹ Fascicle seven of the Abhidharmakośa attacks the same line more fiercely at the end of its section on the four conditions—dharmas do not arise from a single cause such as Īśvara, a self, or primordial nature; if they truly arose from one cause, all things ought to arise at once and not in succession; and it ends with a snort, what use is this Īśvara, and takes apart the name Rudra in passing.² The conclusion is that there is certainly no dharma produced by a single cause.² The first office of a doctrine of causation is not to explain how the world runs; it is to block the road marked only one cause. The quarrel about how many items to have belongs in that setting: the Theravāda and the Sarvāstivāda are doing the same thing, spreading the causing of one event out into many items, and only the manner of spreading differs.

So fewer items does not mean simpler relations. The gap between twenty-four and six-plus-four has not disappeared; it has been pushed elsewhere. Four places are collated below, all of them charged explicitly to this part's account by Chapter 5 of Part 14; where the gap has been pushed will be clear once the four are done.

First, the dominant condition, which bears the same name and nearly the opposite sense. The Sarvāstivāda's dominant condition takes non-obstruction for its meaning: every conditioned dharma, its own substance alone excepted, has all dharmas for its efficient cause, because when it arises nothing stands in its way. The Kośa supplies a simile—the people of a country, because their sovereign does them no harm, all say that it is by the sovereign that they enjoy their ease; and it adds that even a sovereign with no power counts.² Someone presses: if all dharmas are efficient cause to one another, why do not all dharmas arise at once? The answer is that it is only by non-obstruction, and not by any direct productive force upon their arising.² The Theravāda's adhipati-paccaya means the one among desire, mind, energy, and investigation that takes the lead; it is the item that heads a given undertaking, and Part 14 rendered it the predominance condition.⁴ One doctrine says it was not in the way; the other says it went in front. The two are all but contraries.

The point is worth a further sentence. In the Sarvāstivāda apparatus the widest compartment of all is defined by a negative word, and it is precisely that compartment which swallows the efficient cause from among the six. That the widest is the emptiest is not a defect in a taxonomy; it is a choice. To make non-obstruction a compartment is to grant that the great majority of relations need no finer division, and need only be kept from interfering with one another. The Theravāda spends its labour on fine division; the Sarvāstivāda spends its labour elsewhere. Where, the next item shows.

Second, the homogeneous and universal causes against the Theravāda's repetition condition. Both sides are speaking of like producing like in succession, and the reach of the two doctrines is enormously different. The Theravāda's āsevana-paccaya is confined within the seven impulsions of a single cognitive series—the earlier impulsion makes the later one stronger—and outside that series it has no business.⁴ Sarvāstivāda homogeneous production is established across lifetimes: a wholesome root of a former birth is homogeneous cause to a wholesome root of its own sphere in a later birth and to the dharmas associated with it; a past wholesome root is homogeneous cause to future and present ones; a present wholesome root is homogeneous cause to future ones.¹ One is laid out across incalculable lifetimes, the other packed into seven moments.

That it works across lifetimes is possible only on the real existence of the three times, and the treatise says so outright. Fascicle seventeen, on the homogeneous cause, opens as usual by asking why this section is made, and the answer is: in order to stop other schools and display the correct principle; for some hold that past and future have no real substance, some hold that the present is unconditioned, some hold that a dharma is homogeneous cause to its own kind alone—to stop these various divergent holdings and to display that past and future really exist.¹ The first reason for positing the homogeneous cause is to prove that past and future really exist. This is not a connection this book has drawn from outside; it is the treatise's own statement of its legislative motive. Chapter 3 and this chapter are therefore not treating two matters but the two ends of one: the real existence of the three times makes a cross-temporal cause possible, and the cross-temporal cause becomes in turn the most important use to which the real existence of the three times is put. This book takes it to be the most important structural finding of this part—two doctrines propping each other up, either of which pulled away brings the other down at once.

The props hold each other, and the costs pass between them too. Since the homogeneous cause requires before and after, the future stretch runs into trouble: future dharmas lie jumbled without before or after; if they had before and after they would be cause to one another; if cause to one another, then fruit to one another; and mutual cause-and-fruit is contrary to reason.² The conclusion is that there is no homogeneous cause in the future.¹ The bill for that conclusion is not small. Fascicle seventeen has to go through the older texts one by one on its account—the Vijñānakāya, the Prakaraṇapāda, the Prajñaptipāda—and in the end, to the Prajñaptipāda's sentence that dharmas are fixed in four respects, it can only answer that it should be said that in the future world there is no homogeneous cause, while to several other places it answers that they are speaking with a hidden intention, by reference to the last position.¹ To answer with a hidden intention means that those texts, read as they stand, do not agree with this school's position, and must be supposed to mean something else. To save one cause, the treatise re-read three of the six feet.

The same reasoning also keeps the immediately antecedent condition out of the future: future dharmas lie jumbled without before or after, and so there is no immediately antecedent condition in the future.² The question that follows at once is how the Buddha knows what will arise in the next moment. The Kośa records three answers: by inference from past and present dharmas he knows it presently; by the knowledge born of resolve; or, on a third account, there are within the body of a sentient being prior signs of the future fruit and cause, and these are a species of the conditionings dissociated from mind.² The third answer is worth a pause—to patch one sentence, one more dharma dissociated from mind is manufactured. Chapter 4 counted possession, unmanifest matter, the homogeneous character of a group, the life faculty, and the intermediate existence; here is the same move used once more inside the doctrine of causation. Vasubandhu takes none of the three, and says flatly that what the Sautrāntika masters say is best: the Blessed One knows all dharmas by directing his intention, and not by inference or divination.²

This condition carries one more restriction worth recording. The immediately antecedent condition governs mind and mental factors only, and not matter, and the reason is that the quantities do not match: the Bhadanta's examples are that burning a great heap of rice-straw leaves a little ash, which is many producing few, and that one fine seed grows into a nyagrodha tree, which is few producing many.² Mind and mental factors, by contrast, are equal before and after—one mind, one feeling, one perception, neither more nor fewer—and so they alone can be ranged in an unbroken file. Part 14 said that the Theravāda's proximity condition presupposes a single shared clock;⁴ what is presupposed here is that same clock plus a track that carries mind and not matter. Matter's continuity is provided for otherwise, and that is Chapter 6's business. The objective condition runs the opposite way, so wide as to be almost unrestricted: the Kośa says that if a dharma has ever been object to some dharma, there is no time at which it is not object to it, just as firewood is called the burnt even when it is not burning, the characteristic being no different.² Being cognized and being able to be cognized are made one thing, and here again it is the real existence of the three times that pays.

Third, the postnascence condition has no place in this apparatus. The Theravāda posits pacchājāta-paccaya to handle the moment-by-moment mutual support of body and mind: mind arising later supports the material body that arose earlier in its continuing to exist.⁴ Sarvāstivāda causes point uniformly from earlier to later, and there is no such item. On the side of the conditions, however, an opening is left: fascicle fifty-five says that a future view of the existing body, in relation to past and present views of the existing body, is objective condition and dominant condition when taken as object, and dominant condition alone when not.¹ A later dharma can be the object of an earlier one, because future dharmas really exist, and what really exists can be taken as object. So the Sarvāstivāda is not without backward relation; it simply puts every backward relation among the conditions and not one of them among the causes. And this backwardness is asymmetrical—an earlier dharma is dominant condition to a later, and a later is dominant condition to an earlier, yet only the later counts as the dominant fruit of the earlier, and not the reverse.¹ One and the same condition bears fruit when it runs with time and keeps only a title when it runs against.

Fourth, the four items posited for dharmas that have ceased. Decisive support, absence, disappearance, and asynchronous kamma are posited by the Theravāda to answer one question: on what ground can a dharma that has ceased still support a present dharma.⁴ The Sarvāstivāda needs one sentence here—the substance of a dharma exists always; what ceases is the function and not the substance. Part 14's judgement was that the Theravāda pays on the length of the list and the Sarvāstivāda pays in ontology. To audit that account from the Sarvāstivāda side, the question to ask is whether it really paid only once.

It did not. The sentence that the substance of a dharma exists always is not a lump sum but an instalment plan. The first instalment is that the future has to be written as a heap without succession, into which neither the homogeneous cause nor the immediately antecedent condition can enter. The second is that, to preserve the Buddha's omniscience, one more dharma dissociated from mind is added. The third is the heaviest: every cause and every condition must be issued a time of function of its own. Fascicle twenty-one sorts the six causes by time—the associated, co-existent, and causes of maturation run through all three times; the homogeneous and universal causes belong to past and present only; the efficient cause runs through all three times and also stands outside time.¹ A verse of the Kośa pins down the position of the function as well: two causes at the moment of ceasing, three at the moment of arising, and the remaining two conditions exercise their function contrariwise.² The taking of a fruit and the giving of a fruit each have their own times besides: five causes take their fruit in the present only; the associated and co-existent causes give theirs in the present too; the homogeneous and universal causes give theirs in past and present; the cause of maturation gives its fruit in the past alone.² One table of causes and conditions requires a second table of times, and the second is longer than the first.

The four now being done, they may be set out together:

The most naked place in this account is in fascicle sixteen. There the treatise gives each of the six causes a thoroughgoing definition: the associated cause is all mind and mental dharmas; the co-existent cause is all conditioned dharmas; the homogeneous cause is all past and present dharmas; the efficient cause is all dharmas.¹ Fascicle fifty-five says that the four conditions comprise all dharmas—the causal condition comprises all conditioned dharmas; the immediately antecedent condition comprises all past and present mind and mental dharmas except the last mind and mental dharmas of an arhat; the objective and dominant conditions together comprise all dharmas.¹ Both lists rearrange the same heap of dharmas by role, and neither has added one new thing. The holes left are the same two on both sides: a nature is not efficient cause to itself, nor the conditioned to the unconditioned, nor the unconditioned to the unconditioned; the answer given on the four-conditions side is word for word the same.¹² Outside those two places there is not a dharma that can stand outside cause and condition. Yet someone immediately points out that even so the account is not exhaustive, because one still does not know at which position which is cause to which.¹ At which position—those two words are the hinge of the whole chapter: having finished counting the dharmas one has still not stated the causation, and one must further say at which square of time each of them stands.

Three passages have each answered that question: fascicle twenty-one sorting by time, the Kośa verse by the taking and the giving of fruit, fascicle sixteen by the fruit displayed. Laid one upon another they come out thus:

The last column yields only four fruits. The fruit of disjunction has no place on it, and that is exactly why fascicle sixteen stopped counting at four.

Now the fruits. The Theravāda files the fruit under the classification of mind: resultant consciousness is one class of mind divided by function, and its identity is fixed by where it falls in the cognitive series.⁴ The Sarvāstivāda files the fruit under the classification of causal relation: the five fruits are the far side of the causes, not a species of mind. The same matter has been put into different drawers by the two schools, and the choice itself has consequences.

Filed under relation, one must answer how many kinds of fruit there are, and that number was fought over. Fascicle one hundred twenty-one says that the masters of the west posit four further kinds beyond the five—the established fruit, the fruit of preparatory effort, the fruit of conjunction, and the fruit of cultivation—and that the treatise-masters of Kaśmīra say these last four are comprised in the former five, falling under the fruit of human effort and the dominant fruit.¹ The Kośa records the same four and the same subsumption.² Add to this that fascicle sixteen, as noted above, counts only four fruits, and it is plain that five is the outcome of an adjudication and not a fact lying ready to hand. The side that files the fruit under mind need not answer the question at all, because resultant consciousness is a class of mind, and how many classes of mind there are is a separate account.

The fruit of disjunction is the place that will not join up. Cessation through discernment is unconditioned, and the unconditioned has neither cause nor fruit; on what ground, then, is it called a fruit? The Kośa's handling is that the path has a different function with respect to the obtaining and with respect to the cessation: with respect to the obtaining its function is called producing, and with respect to the cessation it is called realizing; the path is not the cause of the cessation, yet cessation through discernment may be said to be the fruit of the path.² The Sautrāntika objects on the spot: the unconditioned is not a cause, and no sūtra says that a cause is unconditioned.² The reply is that although no sūtra says it, no passage forbids it either, and that innumerable sūtras have now vanished from sight—how can one determine that none said it.² The lost scriptures make their second appearance, this time to stop a mouth. This book takes the awkwardness of the fruit of disjunction to lie not in the distinction between realizing and producing but in the fact that this taxonomy has only the one compartment of cause and fruit: the relation between the cultivation of the path and nirvāṇa is not production, and ought properly to have been given a heading of its own; but there are only five headings, and so it was written down as a kind of fruit.

The pair made of the fruit of human effort and the dominant fruit, by contrast, is where this taxonomy does its work well. One and the same finished undertaking is, for the person who did it, fruit of human effort and dominant fruit both; for the person who enjoys it, dominant fruit only. The Mahāvibhāṣā instances husbandry—the crop is fruit of human effort to the farmer and dominant fruit only to the one who eats; the Kośa instances the artisan, one finished article bearing both names for the craftsman who made it and one name only for anybody else.¹² The name of the fruit changes with the end from which you stand, and this is exactly the competence a taxonomy of relations ought to have. A taxonomy of mind cannot state that layer: it can say that this mind is a resultant consciousness, and cannot say whose doing it was.

The converse holds equally. Resultant consciousness in the Theravāda has its determinate place in the cognitive series, and to call it resultant is already to say at which step it appears. The Sarvāstivāda's fruit of maturation can say only that it is exclusively unobstructed-and-indeterminate, and must belong to a sentient being—the receptacle world does not count as a fruit of maturation, since it is held in common, and a fruit of maturation cannot be enjoyed in common.² Each set of terms has its own blind spot. This book's position is that neither is right against the other; there is only the question how the taxonomic choice shapes what can be asked afterwards. Put the fruit inside mind, and the questions run toward the order of the cognitive series; put the fruit inside relation, and the questions run toward how many kinds of fruit there are and whether the unconditioned counts as one.

The last internal strain is in the co-existent cause. The association of mind and mental factors, the mutual co-existence of the four great elements, and mind together with the bodily and vocal action that follows upon it—all three require cause and fruit to be simultaneous; and the criterion is that what are reciprocally fruit of human effort to each other are reciprocally co-existent cause to each other.² A common statement needs correcting in passing: although possession and what is possessed very often arise together, they are not co-existent cause—the Kośa says expressly that possession and what is possessed do not of necessity proceed together, being now before, now after, now co-arisen, and so, although named co-existent, they are not made cause by the co-existent cause.² That exclusion is telling. The possession of Chapter 4 was made precisely so as to be able to fall out of step with what is possessed; the three names prior possession, co-present possession, and posterior possession say just that falling out of step; and the very design that let possessing come loose from the present moment is what disqualifies it from simultaneous causation. Atoms of derived matter that offer resistance are excluded as well, on the ground that they are not of one fruit.²

The associated cause and the co-existent cause are in fact two faces of one thing, and the treatise knows it. Fascicle sixteen asks how the two differ and gives several answers in a row: the sense of being companions is the associated cause and the sense of sharing one fruit is the co-existent cause; the sense of grasping a staff is the associated cause and the sense of having accomplished something with the staff already grasped is the co-existent cause; the sense of joining hands is the associated cause and the sense of having crossed a torrent with hands already joined is the co-existent cause; the sense of according with one another is the associated cause and the sense of being inseparable is the co-existent cause.¹ One and the same batch of mind and mental factors is the associated cause seen as keeping company and the co-existent cause seen as getting one piece of work done together. This is the second time in this chapter that one thing has gone into two drawers, except that this time both drawers are in the Sarvāstivāda's own house. A word in passing: the two shakiest of the six causes are exactly these. The two scriptural sentences assigned to them—that seeing is associated with the faith, realization, and knowledge that are its root, and that eye and form as condition give rise to eye-consciousness, the meeting of the three being contact, together with which feeling, perception, and volition arise—both state that mind and mental factors appear simultaneously, and whether mind and mental factors appear simultaneously is the very thing the Sarvāstivāda has to prove. Adducing it as evidence is circular. The difference of level between the two lists, the dispute between the masters of Kaśmīra and those of the west, and the controversy over the co-existent cause in particular have been sorted out at length in recent scholarship.⁵

Simultaneous causation is wanted because dharmas are too weak. The reason given in fascicle sixteen, in the section on the associated cause, is that conditioned dharmas are feeble in nature and therefore accomplish their work by leaning upon one another; and three similes follow together—two bundles of reeds standing by mutual support, many strands of rope joined able to drag a great log, many men joining hands able to cross a great river.¹ Fascicle fifty-five, on the four conditions, spreads the same thought wider: conditioned dharmas are feeble in their own nature, are not their own masters, depend upon others, have no function of their own, and do not follow their own wish; therefore some arise from four conditions, some from three, some from two, and there is not so much as one that a single condition can produce by itself, still less none; as with a man weak with sickness, four may support him, or three, or two, before he can rise and stand.¹

There is a deep strain here, and it is worth pointing out on its own. Every dharma is really existent and its own nature exists always—that was the conclusion of Chapters 1 and 3; yet the same dharmas, as conditioned, are feeble, without function of their own, not following their own wish. Really existent and powerless, the two statements standing together. This book takes that to be the exact weight of the word own-nature in the Sarvāstivāda: it guarantees only that a dharma has a determinate characteristic of its own, and never guarantees that the dharma can accomplish anything by itself. Part 14 recorded the opposite route in the Theravāda—the force of a condition is intrinsic to the conditioning dharma, as pungency is to a chilli, the relation being drawn back inside one of its own terms.⁴ The Sarvāstivāda cannot take that route, because its dharmas are constitutionally without strength; strength can come only from an arrangement. But the arrangement is not itself a new thing either, since the six causes and four conditions are only the same heap of dharmas rearranged. Strength lies neither inside the dharma nor outside it, and the one remaining place to put it is the square of time on which the dharma stands.

This is where Vasubandhu cuts in. He says that in the causation of seeds and sprouts and the like in the world, simultaneous cause and fruit have never been seen. Someone instances flame and lamplight, or a sprout and its shadow, as simultaneous and reciprocally cause and fruit; he does not accept it—that must first be sorted out: is the flame cause to the light, or did a prior assemblage of causes and conditions make flame and light arise together.² He then produces the logicians' criterion: where, if this is, that is, and if this is not, that is not, this is definitely cause and that definitely fruit.² On that criterion the simultaneous is not necessarily barred from being cause and fruit, but reciprocal cause and fruit will not go through—his own words are that simultaneous cause and fruit may indeed be allowed, but how can one speak of being cause and fruit to each other.² The Sarvāstivāda offers three staves standing by mutual support; he takes the simile apart: are the three staves supporting one another because they were set up together, or did some other cause and condition set them up together—besides which there are ropes and hooks and the ground there, holding them in place.² The reply from the Sarvāstivāda side is that there too there are homogeneous causes and the rest, and that therefore the sense of the co-existent cause stands.²

That reply in fact concedes a step. To rescue a simultaneous cause by means of a prior cause is to grant that the simultaneous layer will not hold up by itself. This book takes the real difficulty to lie not in the simultaneity but in the reciprocity: simultaneous one-way causation is nothing out of the ordinary, and it is being cause to each other that is the awkward thing. The Sarvāstivāda insists on reciprocity because it has already set up characteristic and characterized, and mind and mental factors, as separate substances—that was the work done in Chapters 2 and 4. Things separate in that way must have a cord to bind them; and the moment that cord has a before and an after, a gap opens at the point of binding, a gap this system cannot afford. The co-existent cause is a repair made for a problem that pluralist realism made for itself.

One sentence to close on: the burden of relating is spread by the Theravāda over the items and by the Sarvāstivāda over time. That is not a figure of speech. Fascicle fifty-five, discussing the four conditions, assigns each of them a simile—the causal condition is like a seed, the immediately antecedent condition like clearing a road, the objective condition like leaning on a staff, the dominant condition like not blocking the way—and then says that a later dharma, being taken up by an earlier dharma under the four conditions, is thereby able to proceed through the times, to take its fruit, to perform its work, and to know its object.¹ The first thing purchased by being taken up under the conditions is the ability to proceed through the times, that is, to walk about among the three times. A dharma needs causes and conditions in the first place not in order to arise but in order to be able to move. That sentence settles the whole chapter's account at once: the Sarvāstivāda theory of causation is written in the vocabulary of the real existence of the three times, and cancel that premise and not one term of it remains usable.

To Chapter 6. The standard example of the four great elements as reciprocal co-existent causes has been used repeatedly above, and it was mentioned that atoms of derived matter that offer resistance are barred from reciprocal co-existent causation because they are not of one fruit. Both sentences presuppose a doctrine of matter: what an atom is, how several atoms come to count as one place, and whether the aggregated thing and the atoms gathered in it are the same thing. The Sarvāstivāda works this doctrine out in great detail, and it is here, at the last, that it draws the line between nominal and real existence at a place that would not afterwards hold. The next chapter takes up the material end.

References¶